Re: #include <> vs ""

I did a clean install of OpenBSD 7.7 on new SD card for RaspBerry Pi 4 and #include now works as expected. Thanks again for the help & testing. -KenD

Re: #include <> vs ""

On 2025-07-07 05:07, Stuart Henderson wrote: On 2025/07/06 19:08, ken.dic...@whidbey.com wrote: OpenBSD 7.7 on Raspberry Pi 4 .. Is there anything unusual about how you installed OpenBSD? I can't replicate this on amd64, aarch64, or i386. I am suspecting possible thumb/flash memory failure

Re: #include <> vs ""

On 2025/07/06 19:08, ken.dic...@whidbey.com wrote: > OpenBSD 7.7 on Raspberry Pi 4 > > vvv===vvv===vvv > ~[>> cat trash1.c > #include > #include > int main() { jmp_buf trampoline; exit(1); } > > ~[>> make trash1 > cc -O2 -pipe-o trash1 trash1.c > trash1.c:3:14: error: use of undeclared iden

Re: OpenBSD on ODROID C5

Hi all. Ok, some steps about booting OpoenBSD on the C5: As mentioned, I have a compiled u-boot for the board. I also verify that it works properly and it looks ok. I also have a copy of the dtb, I've loaded it into u-boot and it can print it correctly, so I see it as ok. I got the various me

Re: OpenBSD on ODROID C5

Hi. I was wondering (Yes, I'm on a spending spree and building an additional home lab), the Odroid N2+ has the same processor (S922X) as of the N2. Does anybody had it running OpenBSD? Cheers. -- absc.

Re: OpenBSD on ODROID C5

Hi Mark. On Fri, 23 May 2025 23:41:17 +0200 Mark Kettenis wrote: > > Date: Fri, 23 May 2025 22:00:41 +0200 > > From: Andrea Biscuola > > > > Hi all. > > > > I'm working on trying to get OpenBSD to boot on the new C5. As of > > now, I have a u-boot fimware for the board built. > > > > Given t

Re: OpenBSD on ODROID C5

> Date: Fri, 23 May 2025 22:00:41 +0200 > From: Andrea Biscuola > > Hi all. > > I'm working on trying to get OpenBSD to boot on the new C5. As of > now, I have a u-boot fimware for the board built. > > Given that I'm a newbie into this, advices and suggestions are > appreciated :). > > Anyway,

Re: OpenBSD on ODROID C5

Hi. On Fri, 23 May 2025 22:28:09 +0200 Thomas Dettbarn wrote: > It has been a while, but this is what I did: > > > https://forum.banana-pi.org/t/bpi-r3-how-i-was-able-to-boot-openbsd-7-5-and-failed/18770 > Why, oh WHY the ARM world must be this complicated :D. > Thomas > > On 5/23/25 22:00

Re: OpenBSD on ODROID C5

It has been a while, but this is what I did: https://forum.banana-pi.org/t/bpi-r3-how-i-was-able-to-boot-openbsd-7-5-and-failed/18770 Thomas On 5/23/25 22:00, Andrea Biscuola wrote: Hi all. I'm working on trying to get OpenBSD to boot on the new C5. As of now, I have a u-boot fimware for th

Re: [PATCH 04/10] dwge: add RK3128 support

> From: joshua stein > Date: Wed, 7 May 2025 21:52:24 -0500 > > Before MII reads/writes, make sure it's not busy like the docs say > to. Why? We already have a check at the end of dwge_mii_readreg() and dwge_mii_writereg() to make sure the request completed. So unless we're issuing concurrent

Re: [PATCH 08/10] dwmmc: disable DMA on the RK3128

On Thu, 08 May 2025 at 20:17:11 +0200, Mark Kettenis wrote: > > From: joshua stein > > Date: Wed, 7 May 2025 21:52:28 -0500 > > > > Not sure yet why transfers fail but this gets things going > > I wonder if that's because you don't actually switch the bus into > 4-bit mode? It is now with the

Re: [PATCH 10/10] rkpmic: add rk818 support

> Date: Sat, 10 May 2025 22:43:19 -0500 > From: joshua stein > > On Sat, 10 May 2025 at 14:25:54 +0200, Mark Kettenis wrote: > > > +static const struct rkpmic_regdata rk818_regdata[] = { > > > + { "DCDC_REG1", 0x2f, 0x3f, 0, 0, rk808_vsel_range1 }, > > > + { "DCDC_REG2", 0x33, 0x3f, 0, 0, rk808_v

Re: [PATCH 10/10] rkpmic: add rk818 support

On Sat, 10 May 2025 at 14:25:54 +0200, Mark Kettenis wrote: > > +static const struct rkpmic_regdata rk818_regdata[] = { > > + { "DCDC_REG1", 0x2f, 0x3f, 0, 0, rk808_vsel_range1 }, > > + { "DCDC_REG2", 0x33, 0x3f, 0, 0, rk808_vsel_range1 }, > > + { "DCDC_REG3", 0, 0, 0, 0, rk808_vsel_range1 },

Re: [PATCH 01/10] rkclock: add RK3128 support

> From: joshua stein > Date: Wed, 7 May 2025 21:52:21 -0500 > > Use RK3128_CRU_GLB_SOFTRST_1 to implement cpuresetfn > --- > sys/dev/fdt/rkclock.c| 590 ++- > sys/dev/fdt/rkclock_clocks.h | 52 +++ > 2 files changed, 634 insertions(+), 8 deletions(-) Th

Re: [PATCH 06/10] rkpinctrl: add RK3128 support

> Feedback-ID: i5cf14421:Fastmail > From: joshua stein > > --- > sys/dev/fdt/rkpinctrl.c | 305 +++- > 1 file changed, 303 insertions(+), 2 deletions(-) A few comments on this one. > > diff --git a/sys/dev/fdt/rkpinctrl.c b/sys/dev/fdt/rkpinctrl.c > index 6

Re: [PATCH 10/10] rkpmic: add rk818 support

> From: joshua stein > Date: Wed, 7 May 2025 21:52:30 -0500 > > --- > sys/dev/fdt/rkpmic.c | 27 ++- > 1 file changed, 26 insertions(+), 1 deletion(-) Spotted a mistake... > > diff --git a/sys/dev/fdt/rkpmic.c b/sys/dev/fdt/rkpmic.c > index 0094b61e6b9..ccfa50abbf8 10

Re: [PATCH 07/10] rkusbphy: add RK3128 support

> Feedback-ID: i5cf14421:Fastmail > From: joshua stein > > --- > sys/arch/armv7/conf/GENERIC | 1 + > sys/dev/fdt/rkusbphy.c | 25 + > 2 files changed, 26 insertions(+) ok kettenis@ > diff --git a/sys/arch/armv7/conf/GENERIC b/sys/arch/armv7/conf/GENERIC > index 9

Re: [PATCH 09/10] sdmmc: do 4-bit mode before sending SCR

On Thu, 08 May 2025 at 20:16:33 +0200, Mark Kettenis wrote: > This makes no sense. It is the sdmmc_mem_decode_scr() call above that > sets sf->scr.bus_width. So by removing that call, you're skipping the > bus width change. And I guess that means you'll stick to 1-bit mode? Yeah, ignore this on

Re: [PATCH 08/10] dwmmc: disable DMA on the RK3128

> From: joshua stein > Date: Wed, 7 May 2025 21:52:28 -0500 > > Not sure yet why transfers fail but this gets things going I wonder if that's because you don't actually switch the bus into 4-bit mode? > --- > sys/dev/fdt/dwmmc.c | 5 - > 1 file changed, 4 insertions(+), 1 deletion(-) > >

Re: [PATCH 09/10] sdmmc: do 4-bit mode before sending SCR

> From: joshua stein > Date: Wed, 7 May 2025 21:52:29 -0500 > > Otherwise the dwmmc on the RK3128 fails to "switch func mode 0" > --- > sys/dev/sdmmc/sdmmc_mem.c | 18 +- > 1 file changed, 9 insertions(+), 9 deletions(-) I'm honestly not trying to find an issue with every diff

Re: [PATCH 05/10] mii: add RTL8201F support to rlphy, found on dwge on the XPI-RK3128

> From: joshua stein > Cc: joshua stein > > --- > sys/dev/mii/miidevs | 1 + > sys/dev/mii/miidevs.h | 4 +++- > sys/dev/mii/rlphy.c | 2 ++ > 3 files changed, 6 insertions(+), 1 deletion(-) Two nits below. ok kettenis@ with those fixed. > diff --git a/sys/dev/mii/miidevs b/sys/dev/mii/m

Re: [PATCH 03/10] rkiic: match on RK3128

> Feedback-ID: i5cf14421:Fastmail > From: joshua stein > > --- > sys/dev/fdt/rkiic.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) This diff should not be necessary. The devicetree bindings indicate that the "rockchip,rk3128-i2c" will also list "rockchip,rk3288-i2c" as the compatib

Re: [PATCH 02/10] fdt: add ambrtc, a driver for the Ambiq Artasie AM18X5 RTC

> Date: Thu, 8 May 2025 08:38:31 -0500 > From: joshua stein > > On Thu, 08 May 2025 at 11:52:51 +0200, Mark Kettenis wrote: > > > From: joshua stein > > > Date: Wed, 7 May 2025 21:52:22 -0500 > > > > > > This is found on the XPI-3128 > > > > I'm fairly confident this is the same chip that is

Re: [PATCH 02/10] fdt: add ambrtc, a driver for the Ambiq Artasie AM18X5 RTC

On Thu, 08 May 2025 at 11:52:51 +0200, Mark Kettenis wrote: > > From: joshua stein > > Date: Wed, 7 May 2025 21:52:22 -0500 > > > > This is found on the XPI-3128 > > I'm fairly confident this is the same chip that is already supported > by abcrtc(4). The registers match and the datasheet for t

Re: [PATCH 02/10] fdt: add ambrtc, a driver for the Ambiq Artasie AM18X5 RTC

> From: joshua stein > Date: Wed, 7 May 2025 21:52:22 -0500 > > This is found on the XPI-3128 I'm fairly confident this is the same chip that is already supported by abcrtc(4). The registers match and the datasheet for the Abracon AB18X5 RTC that I can find has a disclaimer that says: AB18X

Re: XPI-3128/RK3128 support

re u-boot, probably best to copy one of the existing u-boot ports to a new directory and modify to use the required upstream. -- Sent from a phone, apologies for poor formatting. On 8 May 2025 03:59:34 joshua stein wrote: These commits get the Geniatech XPI-3128 board working with its

Re: data_abort_handler ci_flush_bp

On Fri, 25 Apr 2025 at 22:24:43 +0200, Mark Kettenis wrote: > > Date: Fri, 25 Apr 2025 19:28:05 + > > From: Miod Vallat > > > > > My RK3128 came up with no ci_flush_bp assigned, because cpu0 never > > > attached, because the reg values in the DTB were 0x000-0x003 rather > > > than 0xf00-0xf

Re: data_abort_handler ci_flush_bp

> Date: Fri, 25 Apr 2025 19:28:05 + > From: Miod Vallat > > > My RK3128 came up with no ci_flush_bp assigned, because cpu0 never > > attached, because the reg values in the DTB were 0x000-0x003 rather > > than 0xf00-0xf03. This caused a "Fatal kernel mode prefetch abort" > > because it wa

Re: data_abort_handler ci_flush_bp

> Date: Fri, 25 Apr 2025 14:25:40 -0500 > From: joshua stein > > My RK3128 came up with no ci_flush_bp assigned, because cpu0 never > attached, because the reg values in the DTB were 0x000-0x003 rather > than 0xf00-0xf03. This caused a "Fatal kernel mode prefetch abort" > because it was tryin

Re: data_abort_handler ci_flush_bp

> > Why not simply skip that call if ci_flush_bp is NULL? > > Because ci_flush_bp should probably be set to something so I figured > the diagnostic is better than skipping it. In my case it was > because mainbus_match_primary() was not matching to the cpu device > when it should have, which pr

Re: data_abort_handler ci_flush_bp

On Fri, 25 Apr 2025 at 19:28:05 +, Miod Vallat wrote: > > My RK3128 came up with no ci_flush_bp assigned, because cpu0 never > > attached, because the reg values in the DTB were 0x000-0x003 rather > > than 0xf00-0xf03. This caused a "Fatal kernel mode prefetch abort" > > because it was tryi

Re: data_abort_handler ci_flush_bp

> My RK3128 came up with no ci_flush_bp assigned, because cpu0 never > attached, because the reg values in the DTB were 0x000-0x003 rather > than 0xf00-0xf03. This caused a "Fatal kernel mode prefetch abort" > because it was trying to execute a null function pointer. > > It took me a while to

Re: rk3128 mmu

> Date: Thu, 17 Apr 2025 17:56:13 -0500 > From: joshua stein > > On Wed, 16 Apr 2025 at 11:54:57 +0200, Mark Kettenis wrote: > > You're accessing the UART registers using their physical address. But > > as soon as you turn on the MMU you need a virtiual address. So you'll > > need to create an

Re: rk3128 mmu

On Wed, 16 Apr 2025 at 11:54:57 +0200, Mark Kettenis wrote: > You're accessing the UART registers using their physical address. But > as soon as you turn on the MMU you need a virtiual address. So you'll > need to create an appropriate MMU mapping for the UART. > > Adding an appropriate entry to

Re: rk3128 mmu

> Date: Tue, 15 Apr 2025 16:26:27 -0500 > From: joshua stein Hi Joshua, > I'm trying to bring up OpenBSD/arm on an RK3128 in the Pomera DM250. > I was able to update its shipped U-Boot from 2014.10 to 2017.09 so > it would have EFI support, and I can boot BOOTARM.EFI. > > It hangs when enter

Re: apple m4 hardware compatibility

Yes, I read some of the Ashai docs and it hardly seems like a trivial task. I appreciate the hard work of all involved. -- ajack...@the-fastest.net On Thu, Apr 10, 2025, at 2:49 AM, Janne Johansson wrote: > Den ons 9 apr. 2025 kl 23:35 skrev : >> >> Nuts. Do you mean that the arm port won'

Re: apple m4 hardware compatibility

Den ons 9 apr. 2025 kl 23:35 skrev : > > Nuts. Do you mean that the arm port won't run on an M4 processor in general? > Why is that, out of curiosity? Running on a new arch or a new device is so much more than just being able to compile code for the cpu. If you can't find the console device, the

Re: RK3528 - Radxa e20c

Hi Mark, I managed to compile a kernel with the patch applied with a bit of fuzz. I put it on the sdcard I had (with snapshot/miniroot77.img + u-boot), but could only fit the non-debug version, so output is missing function names. :-/ I'll see if I can get complete output from debug kernel later

Re: RK3528 - Radxa e20c

Hi mark, On Tue, Apr 8, 2025 at 7:54 PM Mark Kettenis wrote: > So, to start, do you have another OpenBSD/arm64 system on which you > can compile kernels? If so, apply the following patch, build a kernel > and copy that onto the uSD from which you're booting the machine. > > This should print so

Re: apple m4 hardware compatibility

Nuts. Do you mean that the arm port won't run on an M4 processor in general? Why is that, out of curiosity? -- ajack...@the-fastest.net On Tue, Apr 8, 2025, at 3:05 AM, Stuart Henderson wrote: > On 2025/04/07 22:10, ajack...@the-fastest.net wrote: >> If anyone has tried 7.6 or any recent

Re: apple m4 hardware compatibility

the m1n1 bootloader (which we use too) doesn't run on it yet - this is probably the biggest current blocker: https://social.treehouse.systems/@sven/114278224116678776 -- Sent from a phone, apologies for poor formatting. On 8 April 2025 21:35:05 ajack...@the-fastest.net wrote: Nuts. Do you me

Re: RK3528 - Radxa e20c

> From: Vincent Legoll > Date: Sun, 6 Apr 2025 19:12:55 + > > Hi Mark, > > On Sun, Apr 6, 2025 at 7:08 PM Mark Kettenis > wrote: > > We don't have support for the RK3528 SoC yet. In particular there is > no support for this SoC in the rkclock(4) and rkpinctrl(4) drivers > yet. That is

Re: apple m4 hardware compatibility

On 2025/04/07 22:10, ajack...@the-fastest.net wrote: > If anyone has tried 7.6 or any recent version on the newest (2024, apple m4 > cpu) mac mini, please let me know how it worked and of any issues you had > with hardware support. In particular, I'd want to use it with dual monitors > (one on h

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

Le Sun, 6 Apr 2025 22:40:57 +0200, Bernhard Ernst a écrit : > On Sun, 6 Apr 2025 17:46:16 +0200 > wrote: > > > Le Fri, 4 Apr 2025 18:06:08 +0200, > > Bernhard Ernst a écrit : > > > > > Am Fri, 4 Apr 2025 15:24:48 +0200 > > > schrieb : > > > > > > > Le Tue, 1 Apr 2025 11:26:10 +0200, > > >

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

On Sun, 6 Apr 2025 17:46:16 +0200 wrote: > Le Fri, 4 Apr 2025 18:06:08 +0200, > Bernhard Ernst a écrit : > > > Am Fri, 4 Apr 2025 15:24:48 +0200 > > schrieb : > > > > > Le Tue, 1 Apr 2025 11:26:10 +0200, > > > Bernhard Ernst a écrit : > > > > > > > Am Tue, 1 Apr 2025 01:17:45 +0100 > > >

Re: RK3528 - Radxa e20c

Hi Mark, On Sun, Apr 6, 2025 at 7:08 PM Mark Kettenis wrote: > > We don't have support for the RK3528 SoC yet. In particular there is > no support for this SoC in the rkclock(4) and rkpinctrl(4) drivers > yet. That is probably why some of the devices don't work properly. > > Adding support sho

Re: RK3528 - Radxa e20c

> From: Vincent Legoll > Date: Sun, 6 Apr 2025 17:46:29 + Hi Vincent, We don't have support for the RK3528 SoC yet. In particular there is no support for this SoC in the rkclock(4) and rkpinctrl(4) drivers yet. That is probably why some of the devices don't work properly. Adding support s

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

Le Sat, 5 Apr 2025 13:02:03 +0100, Stuart Henderson a écrit : > On 2025/04/04 15:24, b...@courriel.fr.eu.org wrote: > > thx, so no mali drivers under openbsd if I understand correctly (and > > also other GPUs)? > > I _think_ we just have just plain framebuffer drivers for the arm64 > boards, no

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

Le Sat, 05 Apr 2025 06:54:11 -0700, ken.dic...@whidbey.com a écrit : > This won't do for your specs, but Linux in 3 chips is pretty cool! > > > https://dmitry.gr/?r=05.Projects&proj=36.%208pinLinux <3

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

Le Fri, 4 Apr 2025 18:06:08 +0200, Bernhard Ernst a écrit : > Am Fri, 4 Apr 2025 15:24:48 +0200 > schrieb : > > > Le Tue, 1 Apr 2025 11:26:10 +0200, > > Bernhard Ernst a écrit : > > > > > Am Tue, 1 Apr 2025 01:17:45 +0100 > > > schrieb void : > > > > > > > On Mon, Mar 31, 2025 at 04:26:29AM

Re: booting Mac?

OpenBSD in the past and it was reasonably easy to setup. This is unexpectedly difficult. 🙁 * Jay From: Jay K Sent: Saturday, April 5, 2025 1:48 AM To: arm@openbsd.org Subject: Re: booting Mac? I figured out some stuff... the reason the USB drive no longer

Re: booting Mac?

x completely. But if I boot the EFI option, it doesn't get past the early UBoot. Nor does the original flash still boot. I even re-dd'ed it in Linux. It doesn't find its files. I cannot boot OpenBSD is what I mean. Any ideas? Btw, the install is a little unfriendly/unclear. When

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

Le Mon, 31 Mar 2025 07:45:22 -0700, ken.dic...@whidbey.com a écrit : > On 2025-03-30 19:26, b...@courriel.fr.eu.org wrote: > > > I search a board compatible with openbsd for a daily low power > > desktop-mobile computer but I'm kind of lost after too many searches > > and sometimes not enough inf

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

This won't do for your specs, but Linux in 3 chips is pretty cool! https://dmitry.gr/?r=05.Projects&proj=36.%208pinLinux FYI, -KenD

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

On 2025/04/04 15:24, b...@courriel.fr.eu.org wrote: > thx, so no mali drivers under openbsd if I understand correctly (and > also other GPUs)? I _think_ we just have just plain framebuffer drivers for the arm64 boards, no acceleration. > So no 3D, only small video resolution (480p-like) playable

Re: booting Mac?

I can get to the boot menu and back to MacOS, and > even since instaleed Asahi Linux completely. > But if I boot the EFI option, it doesn't get past the early UBoot. > Nor does the original flash still boot. I even re-dd'ed it in Linux. > It doesn't find its files. >

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

I don't own the specific motherboard and I don't know what kind of firmware is running on the RP2040 core, i just assume it isn't seamlessly integrated (eg. it doesn't shows up as a native GPIO device). b...@courriel.fr.eu.org írta 2025. ápr.. 4, P-n 15:24 órakor: > Le Tue, 01 Apr 2025 08:06:27

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

Am Fri, 4 Apr 2025 15:24:48 +0200 schrieb : > Le Tue, 1 Apr 2025 11:26:10 +0200, > Bernhard Ernst a écrit : > > > Am Tue, 1 Apr 2025 01:17:45 +0100 > > schrieb void : > > > > > On Mon, Mar 31, 2025 at 04:26:29AM +0200, b...@courriel.fr.eu.org > > > wrote: > > > >Hello world, > > > >snip<

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

Le Fri, 04 Apr 2025 16:08:56 +0200, MIZSEI Zoltán a écrit : > I don't own the specific motherboard and I don't know what kind of > firmware is running on the RP2040 core, i just assume it isn't > seamlessly integrated (eg. it doesn't shows up as a native GPIO > device). OK, thx

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

Le Tue, 01 Apr 2025 01:16:50 +0200, Kirill A. Korinsky a écrit : > On Mon, 31 Mar 2025 19:07:16 +0200, > Stuart Henderson wrote: > > > > On 2025/03/31 07:45, ken.dic...@whidbey.com wrote: > > > On 2025-03-30 19:26, b...@courriel.fr.eu.org wrote: > > > > > > > I search a board compatible with o

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

Le Tue, 01 Apr 2025 08:06:27 +0200, MIZSEI Zoltán a écrit : > It won't work as a GPIO device under OpenBSD, Why won't it work under OpenBSD? gpio(4) can't play with it? It needs a specific driver for? Is it because of the integrated RP2040 co-processor (and maybe it will work with odroid h4)?

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

Le Mon, 31 Mar 2025 18:07:16 +0100, Stuart Henderson a écrit : > On 2025/03/31 07:45, ken.dic...@whidbey.com wrote: > > On 2025-03-30 19:26, b...@courriel.fr.eu.org wrote: > > > > > I search a board compatible with openbsd for a daily low power > > > desktop-mobile computer but I'm kind of lost

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

Am Tue, 1 Apr 2025 01:17:45 +0100 schrieb void : > On Mon, Mar 31, 2025 at 04:26:29AM +0200, b...@courriel.fr.eu.org > wrote: > >Hello world, > >snip< > > I run OpenBSD 7.6 on a rpi4 (8GB). It's my secondary desktop. My use > case is a little different to yours > > firefox for web > libreoffice fo

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

It won't work as a GPIO device under OpenBSD, but you can get it indirectly working like this: In this case the GPIO is managed by an rp2040 (aka Raspberry Pico 1) chip. Using the Bootsel button on the board you can switch it into LOADER mode, so it shows up as a usb mass storage device. You c

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

On Mon, 31 Mar 2025 19:07:16 +0200, Stuart Henderson wrote: > > On 2025/03/31 07:45, ken.dic...@whidbey.com wrote: > > On 2025-03-30 19:26, b...@courriel.fr.eu.org wrote: > > > > > I search a board compatible with openbsd for a daily low power > > > desktop-mobile computer but I'm kind of lost a

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

On Mon, Mar 31, 2025 at 04:26:29AM +0200, b...@courriel.fr.eu.org wrote: Hello world, snip< I run OpenBSD 7.6 on a rpi4 (8GB). It's my secondary desktop. My use case is a little different to yours firefox for web libreoffice for office stuff mutt for email, but also use a webmail interface No

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

On 2025/03/31 07:45, ken.dic...@whidbey.com wrote: > On 2025-03-30 19:26, b...@courriel.fr.eu.org wrote: > > > I search a board compatible with openbsd for a daily low power > > desktop-mobile computer but I'm kind of lost after too many searches > > and sometimes not enough info. So here I am. >

Re: need help to choose a single-board computer for a daily low power (with good perf) desktop computer

On 2025-03-30 19:26, b...@courriel.fr.eu.org wrote: I search a board compatible with openbsd for a daily low power desktop-mobile computer but I'm kind of lost after too many searches and sometimes not enough info. So here I am. ... At start I was thinking about a RISC-V board but it looks like

Re: aarch64 assembly "hello world"

and with less double-compiling if you put a label at each svc instruction: void x() { __asm volatile( " mov x8, #1;" " _x: svc #0;" ); } typedef unsigned long size_t; int w(void *what, size_t len) { __asm volatile( " mov x2, x1;" " mov x1, x0;" "

Re: aarch64 assembly "hello world"

Den lör 15 mars 2025 kl 09:17 skrev Janne Johansson : > > This might also help: > https://flak.tedunangst.com/post/dude-where-are-your-syscalls > And the text hints at it, but the exact trick to figure out the offsets to put in the "what" ELF section is to compile and link it once, then run $ llv

Re: aarch64 assembly "hello world"

This might also help: https://flak.tedunangst.com/post/dude-where-are-your-syscalls Den fre 14 mars 2025 kl 11:47 skrev Computer Planet : > > Hi guys! > Please, could someone tell me how to print a very simple "Hello, world!" in > assembly for aarch64? > Thanks for reply. > -- May the most sig

Re: aarch64 assembly "hello world"

A working but not simple solution is to make a simple hello.c, and compile it with --save-temps. This means you get a hello.s in there too, and you can build that with cc -O2 -pipe -o hello hello.s and you have a working binary, with the correct ELF sections and everything. It becomes some 5200 b

Re: aarch64 assembly "hello world"

On 2025/03/14 13:20, Alnoman Kamil wrote: > Save as hello.S file > ``` > data > msg: > .ascii"Hello, ARM64!\n" > len = . - msg > .text > > .globl _start > _start: > mov x0, #1 /* fd := STDOUT_FILENO */ > ldr x1, =msg/* buf := msg */ > ldr x2, =len/*

Re: aarch64 assembly "hello world"

Save as hello.S file ``` data msg: .ascii"Hello, ARM64!\n" len = . - msg .text .globl _start _start: mov x0, #1 /* fd := STDOUT_FILENO */ ldr x1, =msg/* buf := msg */ ldr x2, =len/* count := len */ mov w8, #64 /* write is syscall #64 */

Re: Radxa Orion O6

On Thu, 30 Jan 2025 23:53:08 +0100 Mark Kettenis wrote: > > Date: Fri, 31 Jan 2025 00:51:12 +0900 > > From: FUKAUMI Naoki > > > > On 1/30/25 23:59, Patrick Wildt wrote: > > > On Thu, Jan 30, 2025 at 10:07:43PM +0900, FUKAUMI Naoki wrote: > > >> On 1/30/25 19:36, Patrick Wildt wrote: > > >>> Whi

Re: Install on Samsung Book4 Edge with SnapDragon Processor Fails

On 25.2.2025 02:00, tinca...@proton.me wrote: I have a new Samsung Book4 Edge with a SnapDragon Elite X1E84100 3.8 processor. I completed the following steps using an install of OpenBSD 7.6 running in VMware running on a MacBook M1 downloaded the OpenBSD 7.6 ARM64 install76.img and the SHA25

Re: NanoPI r5s

> VICTORY!! > Seems like I have gotten my NanoPI R5S in a state where it can do what I > want it todo. > > --cut-- > pci0 at dwpcie0 > ppb0 at pci0 dev 0 function 0 "Rockchip RK3566" rev 0x00 > pci1 at ppb0 bus 1 > rge0 at pci1 dev 0 function 0 "Realtek RTL8125" rev 0x05: msix, address > 00:00:00:0

Re: Radxa Orion O6

On 1/31/25 07:53, Mark Kettenis wrote: Random process kills occurred on the MP kernel as with the RK3588 board. :( I'm really surprised, because my bootup just works and make -j12 finishes succesfully: 2m42.56s real18m33.46s user 6m20.94s system I've got an Anker 373 120W USB-C

Re: Radxa Orion O6

> Date: Fri, 31 Jan 2025 00:51:12 +0900 > From: FUKAUMI Naoki > > On 1/30/25 23:59, Patrick Wildt wrote: > > On Thu, Jan 30, 2025 at 10:07:43PM +0900, FUKAUMI Naoki wrote: > >> On 1/30/25 19:36, Patrick Wildt wrote: > >>> While I was sleeping Jared made some more progress with NetBSD, see: > >>>

Re: Radxa Orion O6

> Date: Thu, 30 Jan 2025 11:36:38 +0100 > From: Patrick Wildt Hi Patrick, After staring at the ARM ARM for a bit, I don't think the agtimer(4) bits are entirely correct. See below. The xhci(4) bits look good (but I know you're working on a better diff). So the SErrors happen because the hardw

Re: Radxa Orion O6

On 1/30/25 23:59, Patrick Wildt wrote: On Thu, Jan 30, 2025 at 10:07:43PM +0900, FUKAUMI Naoki wrote: On 1/30/25 19:36, Patrick Wildt wrote: While I was sleeping Jared made some more progress with NetBSD, see: * https://mail-index.netbsd.org/source-changes/2025/01/30/msg155355.html * https:/

Re: Radxa Orion O6

On Thu, Jan 30, 2025 at 10:07:43PM +0900, FUKAUMI Naoki wrote: > On 1/30/25 19:36, Patrick Wildt wrote: > > While I was sleeping Jared made some more progress with NetBSD, see: > > > > * https://mail-index.netbsd.org/source-changes/2025/01/30/msg155355.html > > * https://mail-index.netbsd.org/sour

Re: Radxa Orion O6

On 1/30/25 19:36, Patrick Wildt wrote: While I was sleeping Jared made some more progress with NetBSD, see: * https://mail-index.netbsd.org/source-changes/2025/01/30/msg155355.html * https://mail-index.netbsd.org/source-changes/2025/01/30/msg155358.html * https://mail-index.netbsd.org/source-cha

Re: Radxa Orion O6

On Thu, Jan 30, 2025 at 07:55:11AM +0900, FUKAUMI Naoki wrote: > Hi Patrick, > > On 1/30/25 00:55, Patrick Wildt wrote: > > With some input from Jared McNeill I got OpenBSD to install and run on > > the board, it looks like: > > > > * Using xhci(4) produces the SError. Jared thinks it might be m

Re: Radxa Orion O6

Hi, I found a typo. # It doesn't fix our problem. -- FUKAUMI Naoki Radxa Computer (Shenzhen) Co., Ltd. Index: sys/arch/arm64/stand/efiboot/efiacpi.c === RCS file: /cvs/src/sys/arch/arm64/stand/efiboot/efiacpi.c,v diff -u -p -u -r1

Re: Radxa Orion O6

Hi Patrick, On 1/30/25 00:55, Patrick Wildt wrote: With some input from Jared McNeill I got OpenBSD to install and run on the board, it looks like: * Using xhci(4) produces the SError. Jared thinks it might be missing clock setup in UEFI. Disabling it allows it to run userland. It seems

Re: Radxa Orion O6

On Thu, Jan 23, 2025 at 08:49:27AM +0900, FUKAUMI Naoki wrote: > Hi Mark, > > On 1/22/25 22:09, Mark Kettenis wrote: > > I think I figured out what went wrong here. If you know how to apply > > patches and build your own kernel the diff below will probably fix the > > issue. > > It works fine. T

Re: Radxa Orion O6

Hi Mark, On 1/22/25 22:09, Mark Kettenis wrote: I think I figured out what went wrong here. If you know how to apply patches and build your own kernel the diff below will probably fix the issue. It works fine. Thank you very much :) P.S. I think you mentioned GENERIC.MP being unstable on yo

Re: Radxa Orion O6

> Date: Wed, 22 Jan 2025 15:34:45 +0900 > From: FUKAUMI Naoki > > So, what is this panic? It's some sort of hardware error. We've seen this happen when memory is mapped with the wrong memory attributes. Could be the result of the kernel mapping memory that isn't accessable (which in turn could

Re: Radxa Orion O6

> Date: Wed, 22 Jan 2025 14:36:04 +0900 > From: FUKAUMI Naoki > > On 1/22/25 11:49, FUKAUMI Naoki wrote: > > I found the key to enable the SPCR table in EDK2. I rebuilt the UEFI and > > the SPCR table started to appear. But on OpenBSD things got worse, I > > lost output on both serial and HDMI.

Re: Radxa Orion O6

So, what is this panic? (Btw, GENERIC.MP hangs just after "WARNING: CHECK AND RESET THE DATE!") -- FUKAUMI Naoki Radxa Computer (Shenzhen) Co., Ltd. Copyright (c) 1982, 1986, 1989, 1991, 1993 The Regents of the University of California. All rights reserved. Copyright (c) 1995-2025 Open

Re: Radxa Orion O6

On 1/22/25 11:49, FUKAUMI Naoki wrote: I found the key to enable the SPCR table in EDK2. I rebuilt the UEFI and the SPCR table started to appear. But on OpenBSD things got worse, I lost output on both serial and HDMI... ^It was wrong. SPCR is working! [ using 3054328 bytes of bsd ELF symbol t

Re: Radxa Orion O6

Hi Patrick, Mark, On 1/22/25 06:45, Patrick Wildt wrote: If you look at your full dmesg you can see all the tables: [0.00] ACPI: RSDP 0x000479C30018 24 (v02 CIXTEK) [0.00] ACPI: XSDT 0x000479C3FE98 74 (v01 CIXTEK SKY1EDK2 01000101 0113) [0.00] A

Re: Radxa Orion O6

On Wed, Jan 22, 2025 at 05:23:48AM +0900, FUKAUMI Naoki wrote: > On 1/21/25 22:33, FUKAUMI Naoki wrote: > > Hi Mark, > > > > On 1/21/25 07:07, Mark Kettenis wrote: > > > > > > With "ACPI" I can use the serial console with efiboot, but the > > > > > > kernel > > > > > > output only shows up on the

Re: Radxa Orion O6

On 1/21/25 22:33, FUKAUMI Naoki wrote: Hi Mark, On 1/21/25 07:07, Mark Kettenis wrote: With "ACPI" I can use the serial console with efiboot, but the kernel output only shows up on the display after the kernel is loaded. To support the serial console in "ACPI" mode, an SPCR table is needed. P

Re: Radxa Orion O6

Hi Mark, On 1/21/25 07:07, Mark Kettenis wrote: With "ACPI" I can use the serial console with efiboot, but the kernel output only shows up on the display after the kernel is loaded. To support the serial console in "ACPI" mode, an SPCR table is needed. Probably something that should be enabled

Re: Radxa Orion O6

> Date: Mon, 20 Jan 2025 10:25:49 +0900 > From: FUKAUMI Naoki Hi Naoki, > Hi Mark, > > On 1/20/25 00:27, Mark Kettenis wrote: > >> Date: Sun, 19 Jan 2025 12:35:59 +0900 > >> From: FUKAUMI Naoki > > > > Hello, > > > >> Hi, > >> > >> I'm trying to boot OpenBSD/arm64 on a Radxa Orion O6. > >> >

Re: Radxa Orion O6

On 1/20/25 10:19, FUKAUMI Naoki wrote: For now it is not possible to use (serial) console and ukc in ACPI mode, so I compiled a new kernel without them. https://drive.google.com/file/d/11TyxkrAeJN8BtTYkMd02ZWPM1sVf-A88/view? usp=sharing https://drive.google.com/file/d/16yNp0dWgVzUsEZYQGdWoLNEU

Re: Radxa Orion O6

Hi Warner, On 1/20/25 10:36, Warner Losh wrote: On Sun, Jan 19, 2025 at 6:32 PM FUKAUMI Naoki > wrote: Hi Mark, On 1/20/25 00:27, Mark Kettenis wrote: >> Date: Sun, 19 Jan 2025 12:35:59 +0900 >> From: FUKAUMI Naoki mailto:na...@radxa.com>> >

Re: Radxa Orion O6

Hi Patrick, On 1/19/25 23:28, Patrick Wildt wrote: On Sun, Jan 19, 2025 at 12:35:59PM +0900, FUKAUMI Naoki wrote: Hi, I'm trying to boot OpenBSD/arm64 on a Radxa Orion O6. UEFI has a "O/S Hardware Description Selection" menu where I can choose "Device Tree" or "ACPI". "ACPI" allows me to boot

  1   2   3   4   5   6   7   8   9   10   >